The present invention relates to medical procedures that are performed with reference to images of the patient and, more particularly, to medical procedures performed with reference to projective images such as fluoroscope images, and also to medical procedures performed with reference to images acquired prior to, and independently of, the procedures.
Images of the interiors of patients commonly arc used to guide the performance of invasive medical procedures on the patients. Bucholtz, in U.S. Pat. No. 5,383,454, Ferre et al., in U.S. Pat. Nos. 5,829,444 and 5,873,822, and Bourman, in U.S. Pat. No. 5,902,239, teach the navigation of a probe, such as a catheter, within the body of a patient, with reference to previously acquired images. Barrick, in U.S. Pat. No. 5,772,594, teaches fluoroscopic imaging of a bone prior to the insertion therein of a guide pin or screw with reference to the image.
Two kinds of imaging modalities are in common use. Representational images, such as CT images, MR images and ultrasound images, represent physical properties of the patient's body at particular locations therein. For example, each pixel of a 2D digital ultrasound image of a patient represents an acoustic impedance contrast at a corresponding point inside the patient's body, and each voxel of a 3D CT image volume represents the density of the patient's body tissue at a corresponding point inside the patient's body. Projective images, such as fluoroscopic X-ray images, represent projections of physical properties of the patient's body into a plane. For example, each point in a fluoroscopic X-ray image is an integral along a ray, from the X-ray source to the X-ray image, of the density of the patient's body tissue.
Two prior art references of particular note are Gilboa et al., WO 00/10456 and WO 00/16684, both of which documents are incorporated by reference for all purposes as if fully set forth herein. WO 00/10456 teaches intra-body navigation of a probe in conjunction with imaging by a C-mount fluoroscope. WO 00/16684 teaches the use of a representational imaging device, such as an ultrasound probe, in conjunction with the C-mount fluoroscope of WO 00/10456, for the purpose of identifying and recording points-of-interest, within the body of the patient, towards which the probe subsequently is navigated. FIG. 1A, which is adapted from FIG. 2 of WO 00/16684, shows a patient 24 lying on an operation platform 34 and being imaged by a C-mount fluoroscope 22. A catheter 26 is navigated within a body cavity 28 of patient 24. This navigation is enabled by the provision of a transmitter 30 of electromagnetic radiation under-platform 34, a receiver 40 of electromagnetic radiation rigidly attached to fluoroscope 22, and a receiver 32 of radiation rigidly attached to catheter 26, all three of which are connected by wires 51 to a computer 50. Fluoroscope 22 is used to acquire an image of a portion of the body of patient 24 that includes body cavity 28. As explained in WO 00/10456, transmitter 30 defines a reference frame, and the signals received by computer 50 from receivers 40 and 32 in response to the electromagnetic radiation transmitted by transmitter 30 are indicative of the locations and orientations of fluoroscope 22 and catheter 26 relative to the reference frame. Given these locations and orientations, computer 50 displays, on a display unit 48, the image of body cavity 28 acquired by fluoroscope 22, with an icon representing catheter 26 superposed on the image in the true location and orientation of catheter 26 relative to body cavity 28.
Because patient 24 may move, relative to platform 34, during the medical procedure, patient 24 also is provided with a receiver 38 of electromagnetic radiation. Computer 50 computes, from the signals received from receiver 38 in response to the electromagnetic radiation transmitted by transmitter 30, the location and orientation of the body of patient 24 relative to the reference frame. This is in addition to the computation, by computer 50, from the signals received from receiver 40, of the location and orientation of fluoroscope 22 relative to the reference frame, and in addition to the computation, by computer 50, from the signals received from receiver 32, of the location and orientation of catheter 26 relative to the reference frame. Computer 50 records the location and orientation of patient 24 when the image of body cavity 28 is acquired. If patient 24 does moves, computer 50 adjusts the joint display of the image and the catheter icon on display unit 48 to reflect the movement of patient 24, so that the catheter icon always is displayed in a manner that reflects the true location and orientation of catheter 26 relative to body cavity 28.
As alternatives to receivers 32 and 44, catheter 26 and patient 24 are provided with respective imagable markers 46 and 44. Computer 50 locates the shadows of markers 46 and 44 in the image, using standard image processing techniques, and computes, from the locations of these shadows, the locations and orientations of catheter 26 and patient 24.
A representational imaging device 52, equipped with a receiver 40a of electromagnetic radiation, also is provided, to acquire a representational image of a portion of the body of patient 24 that overlaps with the portion of the body of patient 24 that is acquired using fluoroscope 22. Computer 50 computes, from the signals received from receiver 40a in response to the electromagnetic radiation transmitted by transmitter 30, the location and orientation of imaging device 52 relative to the reference frame. Computer 50 then displays the representational image, on display unit 48, superposed on the image acquired by fluoroscope 22, so that a point-of-interest, towards which catheter 26 is to be navigated, can be picked on display unit 48, even prior to the introduction of catheter 26 into body cavity 28. Improved methods of effecting this superposition are taught by Gilboa et al. in PCT application US99/26826, which also is incorporated by reference for all purposes as if fully set forth herein. Because computer 50 tracks the movement of both patient 24 and catheter 26, an icon representing the point-of-interest is displayed on display unit 48 in a manner that represents the true location of the point-of-interest in body cavity 28, so that catheter 26 can be navigated Jo the point-of-interest with reference to the relative locations, as displayed by display unit 48, of the icon representing catheter 26 and of the icon representing the point-of-interest.
Representational imaging device 52 may be external to the body of patient 24, as illustrated in FIG. 1A, or internal to the body of patient 24. The specific example of representational imaging device 52 that is presented in WO 00/16684 is an intracardiac ultrasound probe that is used to image and identify the fossa ovalis of the cardiac septum and one or more of the openings of pulmonary veins. These points within the heart may be targets of ablation for treating atrial fibrillation, and so constitute points-of-interest within the heart (as body cavity 28) of patient 24. Following the representational imaging of these targets by intracardiac ultrasound probe 52 and the picking of the points-of-interest, intracardiac ultrasound probe 52 is withdrawn and an ablating catheter 26 is navigated towards the points-of-interest.
Transmitter 30 is an example of what is called in WO 00/16684 a “locating implement”. Receivers 32, 38, 40 and 40a are examples of what is called in WO 00/16684 “location implements”. Transmitter 30, together with receiver 32, 38, 40 or 40a constitute what is called in WO 00/16684 a “locating system”. In the present context, the location and the orientation of an object are called the “disposition” of the object, so what WO 00/16684 calls a “locating system” is called herein a “disposing system”. Similarly, what WO 00/16684 calls a “locating implement” is called herein a “disposing implement”, and what WO 00/16684 calls a “location implement” is called herein a “disposition implement”. The term “location system” is used herein to refer, not only to systems that measure both the location and the orientation of an object, but also to systems that measure only the location of an object; such a system includes a locating implement and a location implement. Note that when multiple disposing systems are used, one disposing instrument may be shared by all the disposing systems as is the case with transmitter 30, in which case the shared disposing instrument defines a common reference frame for all the disposing systems; or, alternatively, one disposition implement may be shared by all the disposing systems, in which case the shared disposition instrument defines a common reference frame for all the disposing systems. Common examples of disposing systems include electromagnetic disposing systems, magnetic disposing systems, acoustic disposing systems and stereopair optical systems.
One invasive medical procedure for which the prior art methods are not quite suitable is the deployment of a stent in a partially blocked coronary artery. This procedure commonly is performed by injecting a contrast agent into the target coronary artery tree and then navigating a catheter that bears the stent towards the target blockage with the help of X-ray angiographic images acquired in real time by a fluoroscope such as fluoroscope 22. In this case, the points-of-interest are the blockage itself, and the branches of the coronary artery tree that must be traversed on the way to the blockage. In principle, the prior art discussed above can be used to identify the points-of-interest, provided that these points-of-interest can be picked on the image provided by representational imaging device 52. For example, representational imaging device 52 may be a CT scanner: the contrast agent, being X-ray opaque, shows up in both the projective images acquired using fluoroscope 22 and the representational CT scan acquired using the CT scanner. This has the disadvantage of requiring the use of two imaging modalities, one of which (CT) is not suitable for real-time imaging. Furthermore, it is relatively difficult to register a CT image volume with a fluoroscopic image.
There is thus a widely recognized need for, and it would be highly advantageous to have, a method, of navigating a probe to a point-of-interest in a body cavity of a patient, that is based on a single projective imaging modality.
A CT scanner produces its representational image volume by appropriate processing (typically, by backprojection) of a set of projective images. In principle, then, it should be possible to use fluoroscope 22 itself as both a projective imager and a representational imager. Yeung, in U.S. Pat. No. 5,588,033, which is incorporated by reference for all purposes as if fully set forth herein, teaches the reconstruction of a binary (two-level) image volume from a relatively small set of projective radiographic images. In principle, a similar reconstruction should be possible using fluoroscopic images acquired at different dispositions relative to patient 24. In particular, in the stent deployment discussed above, a binary representational image volume of the contrast agent in the coronary artery tree would show the portion of the coronary artery tree that contains the contrast agent at one of the two display levels (e.g., “1”) and the rest of the imaged portion of the patient's body at the other level (e.g., “0”), and so would suffice to allow picking of the points-of-interest. In practice, however, fluoroscope 22 lacks sufficient mechanical stability to allow accurate reconstruction of even a binary image volume. Known successful reconstructions of image volumes from 2D projective images all require very accurate positioning of the imaging apparatus. Conventional CT scanners include heavy and very accurate mechanisms for rotating their X-ray sources and detectors and similarly heavy and accurate sliding mechanisms for moving the platform on which the patient lies. Yeung uses a stereotactic localizer frame to provide the required dispositional accuracy.
There is thus a widely recognized need for, and it would be highly advantageous to have, a method of transforming a set of projective images, acquired using a projective imager of limited mechanical stability, into a representational image volume.
Returning to the procedure for deploying a stent in a coronary artery, the fluoroscope commonly is placed in a disposition relative to the patient that is expected to give the best projective view of the target coronary artery tree. The contrast agent is injected into the coronary tree, and the projective image is acquired and digitized. This projective image is used as a background “road map” for catheter navigation with the help of other images subsequently acquired of the target coronary artery tree, but only if the fluoroscope and the patient remain in the same relative disposition. Movement of either the fluoroscope or the patient renders this projective image useless as a road map. In particular, if the disposition of the fluoroscope relative to the patient turns out to be suboptimal, or if the patient must be imaged from several dispositions of the fluoroscope in order to give an adequate picture of the three-dimensional structure of the coronary artery tree, then for each new disposition of the fluoroscope, the contrast agent must be injected anew and a new road map must be acquired. This exposes both the medical team and the patient to additional X-radiation, and also exposes the patient to the danger of liver damage from repeated injections of the contrast agent.
There is thus a widely recognized need for, and it would be highly advantageous to have, a method of acquiring and using X-ray angiographic road maps without undue danger to either the patient or the medical team.